Scoring the teachers
August 27, 2010 3 Comments
The chart above is from a study done by the LA Times on school teachers. It essentially tracks the performance of a class on standardised math and reading tests over multiple years and correlates that to the teacher involved. Since the class remains reasonably consistent over a year, the analysis takes care of sample bias issues that normally creep into teacher evaluations.
As one can imagine, the results have caused quite a debate. The teachers union doesn’t like it (why would they!) and the parents have gotten quite curious about who they’re sending their kids to. While the dust is still settling, it seems that some form of this assessment is going to become ingrained in the LA teaching structure – and that a lot of the US may end up going this way.
There are obviously merits to this approach. Parents, tax-payers, government have all got a right to understand just how good the teachers are, or aren’t. But by focusing on standardised test scores (English and Math only), one can also cause teachers to teach-to-a-test – something we’ve all seen a lot of in India.
India struggles with coverage, the US with quality, Korea with the cram factories. No one’s happy, which means there’s a lot of opportunity!
amtgrg
Gaurav Rastogi
Two MBAs, One Blog 
Performance measurement of teachers can be a tricky thing. One of the first thing that starts happening is that the organization or the individual starts screening the entrants. There have been studies in US schools which showed that in schools where teacher/school performance was measured, the school stopped trying to reduce drop-out rates. So they had fewer “problem” children. This can be seen in all mainstream schools in India and even tuition centers which like to talk about what % of their kids reached what ranks… And isn’t it the same in corporate settings: If people are measured simply against goal achievement, they try to set low goals.
Oh absolutely. Which is why this study is pretty good – it adjusts for the starting level of the students in the class and measures change over time. So if a kid starts with 40% and gets to 50% thats better than someone starting at 80% and getting to 83%. Intuitively, thats the way a parent would view it too.
Pingback: Incentives to learn « Two MBAs, One Blog